Thanks, appreciate these insights. I've been contemplating VA's Oracle / Cerner implementation and wondered if implementing the same systems across…
EHR Design Talk with Dr. Rick 2/6/12
Why T-Sheets Work
Disclosure: I have no financial interest in T-System, Inc.
There is nothing particularly high-tech about a T-Sheet. A T-Sheet (designed by T-System, Inc.) is a particular design for a double-sided, single-page printed paper form used to chart patient visits. T-Sheets are extremely popular and have been widely adopted by emergency department and urgent care physicians.
Why do many physicians prefer using T-sheets to the more technologically advanced EHR solutions that they are increasingly being required to adopt?
There are of course many reasons. One is so basic — and is such a defining property of the paper form in general — that we tend not to even notice it: T-Sheets assign each category of data to a box of fixed size and fixed location on the page.
A second reason T-Sheets are popular is that each presenting problem (chest pain, abdominal pain, headache, and so forth) has its own customized T-Sheet template. But regardless of the specific problem and the specific data collected, the spatial layout of data categories is kept exactly the same.
Here is an example of the front side of a T-Sheet for an emergency department visit that I have redrawn and greatly simplified to emphasize its high-level spatial design.
Regardless of the reason for the emergency department visit (in this case, chest pain), the box on the top right has a fixed size and location. It is always set aside for the review of systems (ROS). Similarly, regardless of the reason for the visit, the box on the bottom right has a fixed size and location. It is set aside for the family history. And so forth.
This means that once I learn where each category of data is situated on the page, I can just glance at that box to retrieve the desired information. Its position doesn’t change depending on how much data is written in the boxes above or next to it. The information remains readily available when I’m viewing a different box. I don’t have to carry it in my head.
The locations become automatic after a while. I don’t have to read the box headings. And if I need to compare the current visit to a previous one, I can just place the two T-Sheets side-by-side and glance at the same location on the two sheets to find the comparable data.
In my last post, Computer-Centered versus User-Centered Design, we saw how the spatial arrangement of data allows us to solve certain problems visually with minimal cognitive effort. But even if our task is just to take in and organize a large amount of data, a fixed spatial arrangement is a very good design.
Humans are visual animals par excellence. The human visual system is very good at organizing objects in space. T-Sheets and similar paper forms work because they enable us to use our extraordinary visual and spatial processing abilities to make sense of abstract data, even though these abilities evolved to help us organize physical objects in the real world.
Despite its simplicity, the paper form — with every data category assigned to a fixed location on the page — is a powerful cognitive tool. By allowing us to use our perceptual visual system to organize and retrieve a large body of information, it leaves our finite cognitive resources available for patient issues.
This all may seem obvious. Unfortunately, many EHR designs did not go in this direction, only in part because of technical constraints. Instead, clinicians often are required to navigate to multiple screens in order to enter or view different categories of data, as in the example below:
Of course paper forms have their own problems — how do you record more information than fits in a particular box, bring historical information forward to the next encounter without laboriously re-entering it, read illegible handwriting, and so forth? But still, assigning each data category a fixed screen location is a good model. So in rethinking EHR design, one strategy is to retain fixed spatial location as a high-level design element, but improve the paper design by making it interactive.
We need interactive T-Sheets.
Next Post:
Humans Have Limited Working Memory
Rick Weinhaus MD practices clinical ophthalmology in the Boston area. He trained at Harvard Medical School, The Massachusetts Eye and Ear Infirmary, and the Neuroscience Unit of the Schepens Eye Research Institute. He writes on how to design simple, powerful, elegant user interfaces for electronic health records (EHRs) by applying our understanding of human perception and cognition. He welcomes your comments and thoughts on this post and on EHR usability issues.
Good news, Dr. Weinhaus. You can buy them right here
http://www.tsystem.com/
And they are Meaningful Use certified now too…
There are interactive T-sheets, the T-SystemEV. Unfortunately it doesn’t seem to translate all that well to the screen. What is systematic on paper is kind of chaotic on screen. I don’t recall if they kept the information collection spatially uniform or not.
Hey, I recognize that Balsamiq screen pic! Great little program.
The paper T-System (and other good chief complaint based PAPER template systems) works because it is the most efficient and best way to meet the charting needs of the ED doc. Good paper template sytems are the preferred method of charting for far greater than 90% of ED docs
ED doc charting needs:
1. Get paid: this was of course the reason the T-system started was as a billing and coding tool. Since most ED groups are billed/coded by a 3rd party company who are petrified of an audit, it is easy to ‘check the boxes’ to meet the requirements. Furthermore, a coder can look at the document and in about 2 seconds and determine ‘all boxes checked, cha-ching, level 5’
2. Being able to figure out what you were thinking when someone wants to throw stones at you (lawyers, QA people, etc.): by writing in the margins and free texting, it’s fairly easy to do with a paper template. Computer templates produce garbage charts with no useful distinguishing info unless you type a lot.
3. Be efficient: paper templates are done at the point of care and TIME of care; charting usually done while you are at the bedside. Not only is this the most efficient, but you are also spending more time with the patient.
Computer templates are NOT AT ALL LIKE paper templates, the reason being that mouse and keyboard data entry cannot match the mobility, efficiency, and ‘usability’ of a pen (the electronic T-System is garbage as well). This is very poorly understood, and virtually nothing is written about this, but my 2 favorite pieces are a short paper:
http://www.ama-assn.org/amednews/2003/02/17/bica0217.htm
… and a facscinating short little book, “The Myth of the Paperless Office” by Sellen and Harper
I am surprised at the comments regarding T-EV. I’ve been a user of the system since 2006 in my ED. I came from a paper-based dictation system followed by T-Sheets before installing T-EV. There is no question that T-EV has resulted in increased reimbursement. Although I have always thought that any electronic charting system would have an efficiency hit, that was not the case when we went to T-EV. What we measured was a decline in patient arrival to doctor, reduced overall ED stay, and reduced left without being seen. These changes were seen in the months after installation and have persisted to this day. Additionally, where we were spending up to 2 hours each shift catching up on paperwork, docs (and nurses) get out on time with charts completed! That is a huge satisfier as I’m sure you’d appreciate.
I absolutely agree that most systems are painful and get in the way of patient care in the ED. But T-EV is clearly not like other systems in that regard. There is no greater satisfier of patients than to have me sitting at the bedside with my tablet computer documenting as I take the history and writing orders via CPOE that get started often while I’m still in the room doing the exam and history.
I think all of us in medicine must hold vendors to their promises. They all promise the stars. But, few actually deliver. But it is clearly not accurate to paint all of them with the same brush as much as we would like to sometimes. I for one, would seriously consider not taking a job somewhere if they did not have T-EV in the ED. I’ve looked at a bunch of other systems and it would ruin my day to be subjected to the pain and suffering of other badly implemented electronic systems. For me T-EV has been a terrific solution.
Re: “We need interactive T-Sheets.”
An interactive T-Sheet is available, and the ED where I practice has been using it with great success for the last 18 months as a beta test site. The product is called DigitalShare from T-System, which is powered by Shareable Ink technology. (This system covers all of the “meaningful use” criteria.) Prior to implementing this system, we had been using paper T-Sheets and a regular pen for our documentation. While we continue to document using the paper T-Sheets, the PEN we use has changed. Shareable Ink’s computerized pen is used to write – in ink – on a paper T-sheet as if the pen is a regular ink pen. When the pen is placed in the electronic dock, the data is automatically transferred into the EHR. Because the pen was the only thing that changed for our documentation, the transition was hardly noticeable (even for some of us docs who are somewhat computer illiterate). In fact, our ED, which sees 2.7 patients/hour, has maintained its satisfaction rate of greater than 90%.
I am so satisfied with this system that I highly recommend all ED directors check it out for themselves. I always leave the ED at the end of my shift with all of my documentation complete.
Note: I have no financial interest in T-System, Inc. or Shareable Ink (but I think I might look into it!)
Margalit, Brian, Lizzie, disbelieving doc, Dean Gushee, and Dr. Dan — Thanks for your comments!
In one way or another, all of you touch on the same issue — namely — what happens when you take a mechanical-age cognitive tool (such as the paper form) and try to represent it using digital-age technology?
Alan Cooper has a great discussion of this problem in About Face 3, the Essentials of Interaction Design. He makes a strong case that new technologies demand new representations. He writes that when cognitive tools are brought over unchanged from the mechanical age to the digital age, they tend to combine the weaknesses of the old with the weaknesses of the new.
Compared to the long history of designs for the medium of pen and paper, the interface designs for EHRs are still in their early stages. For EHR designs to function effectively as cognitive tools, they need to take advantage of the unique strengths of digital technology, which are very different from the strengths of pen and paper.
It is my hope that these posts and the discussion generated by them will contribute to the development of new EHR designs that better support the clinicians who use them.
Rick
I’m commenting as a long time reader of HIStalk and a parent whose 19 year had to go to the ER within 24hrs of a previous ER visit. When the ER doc on the second visit tried to determine what was done by looking at the faxed T-sheet they couldn’t. No one could read the scribble on the eval sheet. So while I’m sure a physician would say writing is “easier and faster and mayber so long as I get paid”, there is no reason with the technology that’s available that these manual systems should still be out there.
Mark — Thanks for your comment reinforcing the fact that paper documention has serious deficiencies, illegibility being high on the list. There is no doubt that we are moving inexorably toward digital solutions. The question is how to design these digital user interfaces so that they are, on the whole, better than the paper technologies that they are replacing.
Rick
I am confused. While agreeing wholeheartedly with your underlying premise, I disagree with your example.
I have documented many thousands of emergency visits using T-sheets so I am quite familiar with them. You describe one of the greatest ‘virtues’ of the paper T-sheets as a consistency of layout, with the same information occupying the same location on the sheet irrespective of the T-sheet chosen. There are scores of different T-sheets for different problems so a consistent layout would indeed be a virtue, lowering the hurdle of learning scores of different forms.
The problem is that your description is completely wrong!
Every T-sheet template has a unique layout. Sometimes the ROSS is in bottom area of the first column, other times it is the top of the second column. Sometimes the physical exam starts at the top of the second column and finishes at the bottom; other times it starts somewhere else and ends on the back of the sheet. The ORDER of the sections is consistent, but after that content simply flows and wraps.
Further, looking to the details under the various headings (e.g. ROSS, PMH, PE, etc.) different T-sheets (i.e.different problems) prompt for different ‘sets’ of data. The SAME PMH item will be in a DIFFERENT location (or not present at all) on different T-sheets. Age and sex differences are not at all addressed, other that having some pediatric-specific templates.
The net of this intersection of layout and content is that each T-sheet is a unique form and unless one uses the forms for YEARS, they will not develop an experience and a ‘muscle memory’ for anything other than an undiscerning and indiscriminate ‘slash-em-all’ in order to meet reimbursement requirements. Confound this with inadequate room to write patient specific detail and treatment/response/medical-decision-making and you end up with a complete mess.
While first agreeing the consistent layout of information would dramatically help in the efficiency of use of paper (and electronic forms) I will assert that paper T-systems is NOT an example of such a strategy. Further, the fundamentally flawed notion that the appropriate care and documentation of an emergency visit is inexorably linked (by the choice of a paper template record AND/OR most electronic charts) to the first impression of the nature of the patient’s presenting chief complaint is WRONG and risks quite serious logistical, operational, and cognitive dysfunction as the patients’ care proceeds.
While I would not go so far as to blame the T-sheet forms, the overall quality of records (measured in information transmission as opposed to whether they are billable) generated with paper T-sheets is among the worst that I have experienced in 30 years of practice. While I have not used T-Systems EV (electronic charting), my limited experience with READING the output of that system would put these toward the opposite/superior end of the quality spectrum.
Restating your opening query: “Why do many physicians prefer using T-sheets to the more technologically advanced EHR solutions that they are increasingly being required to adopt?”
My answer: They are the fastest and simplest (albeit among the least informative) mechanisms available to generate a chart that will result in the highest, most appropriate and most-likely-to-be-paid bill possible. When you are seeing 2.5 to 3.5 patients per hour that is critical.
Charles — thanks so much for your post!
First of all, let me state that I have no first-hand experience with T-Sheets and I certainly defer to your years of personal experience using them.
My information about the consistent layout of data categories comes from my phone and e-mail conversations with T-System, Inc. They sent me ED T-Sheets for several different presenting complaints and also affirmed that there was a consistent layout of data categories for all medical ED T-Sheets. The T-Sheets I studied did in fact stick to a consistent layout for the data categories. As you observe, T-Sheets do vary in terms of the specific content within those categories, but this seems to me to be a necessity and does not detract from my main point.
T-System also informed me that their trauma ED T-Sheets have a different but consistent layout. For the purposes of my post, I did not go into this level of detail.
So, in good faith, when I chose T-Sheets as an example of the paper patient visit form, I was under the impression that ED T-Sheets for medical problems all had a consistent layout on the page for the major data categories. I appreciate your input very much. If all medical ED T-Sheets do not have a consistent layout, then I stand corrected.
As you know, the real purpose of my post was to advocate for a consistent spatial layout as a high-level design principle when documenting a patient visit. I happened to choose T-Sheets because they are a well known example of the printed paper form. In my posts, I’m trying to include as many concrete examples as possible, instead of just writing in the abstract. On the basis of your input, it sounds like T-Sheets may not have been the best example to use.
As to your second point, I wholeheartedly agree that any design (paper-based or electronic) in which the first impression of the patient’s problem easily determines the rest of the evaluation is fraught with risks to patient care and safety. One such risk is “anchoring,” where all subsequent information that the clinician collects, even if contradictory, gets forced into the framework consistent with the initial incorrect assessment. As you know, these template-based designs are ubiquitous and inhibit the kind of open-ended listening and questioning necessary for good patient care.
And yes, you are right that T-Sheets, like almost all EHR solutions, often require the clinician to document irrelevant clinical information just for billing purposes. I’m not sure that the vendors are driving this — see Mr. H’s poll in this week’s Monday Morning Update.
So thanks again both for your helpful information and for your thoughtful comments. I look forward to your continued input.
Rick